1 \documentclass[ignorenonframetext]{beamer}
3 \setmainfont{FreeSerif}
9 \usepackage[bf]{caption}
14 % \usepackage{multirow}
17 \usepackage[backend=biber,natbib=true,hyperref=true,style=nature]{biblatex}
18 \addbibresource{references.bib}
19 % \usepackage[nomargin,inline,draft]{fixme}
20 % \newcommand{\DLA}[1]{\textcolor{red}{\fxnote{DLA: #1}}}
21 % \usepackage[hyperfigures,bookmarks,colorlinks,citecolor=black,filecolor=black,linkcolor=black,urlcolor=black]{hyperref}
25 \usepackage{zref-xr,zref-user}
26 \renewcommand*{\bibfont}{\tiny}
28 % The textpos package is necessary to position textblocks at arbitary
29 % places on the page. Use showboxes option to show outlines of textboxes.
30 % \usepackage[absolute]{textpos}
31 \usepackage[absolute,overlay]{textpos}
32 \usepackage{mathtools,cancel}
34 \renewcommand{\CancelColor}{\color{red}} %change cancel color to red
40 \usepackage{tcolorbox}
42 \BeforeBeginEnvironment{minted}{\begin{tcolorbox}}%
43 \AfterEndEnvironment{minted}{\end{tcolorbox}}%
46 \usetheme{CambridgeUS}
48 % http://identitystandards.illinois.edu/graphicstandardsmanual/generalguidelines/colors.html
49 \definecolor{ilboldblue}{HTML}{002058}
50 \definecolor{ilboldorange}{HTML}{E87722}
51 \definecolor{ilblue}{HTML}{606EB2}
52 \definecolor{ilorange}{HTML}{D45D00}
53 \logo{\begin{tikzpicture}% Pale figure
54 {\node[opacity=0.1]{\IfFileExists{figures/uofi_mark.pdf}{\includegraphics[width=2cm,height=1cm,keepaspectratio]{figures/uofi_mark}}{}%
59 % remove navigation symbols
60 \setbeamertemplate{navigation symbols}{}
62 \title[Simons Genome Div.]{Simons Genome Diversity Project}
63 \author[Don Armstrong]{Don L. Armstrong}
64 \institute[IGB]{Institute for Genomic Biology, Computing Genomes
65 for Reproductive Health, University of Illinois, Urbana-Champaign}
69 <<load.libraries,echo=FALSE,results="hide",warning=FALSE,message=FALSE,error=FALSE,cache=FALSE>>=
70 opts_chunk$set(dev="cairo_pdf",out.width="\\textwidth",out.height="0.8\\textheight",out.extra="keepaspectratio")
71 #opts_chunk$set(cache=TRUE, autodep=TRUE)
72 options(device = function(file, width = 8, height = 7, ...) {
73 cairo_pdf(tempfile(), width = width, height = height, ...)
84 \IfFileExists{figures/relevant_xkcd.png}{\frame[plain]{\centering \includegraphics[width=\textwidth,height=0.8\textheight,keepaspectratio]{figures/relevant_xkcd.png}
86 \url{https://xkcd.com/1706/}}}
88 \frame[plain]{\titlepage
90 Code and slides are here:
92 \qrcode[padding]{http://dla2.us/p/genomdiv2016}
94 \url{http://dla2.us/p/genomdiv2016}
101 \includegraphics[width=\textwidth,height=0.8\textheight,keepaspectratio]{genome_diversity_paper/title.png}
105 \begin{frame}{Sampling}
107 \item 142 populations from Africa, America, Oceania, South Asia,
108 East Asia, and West Eurasia (mostly indigenous)
109 \item 300 samples sequenced at 34-83 fold coverage by Illumina
110 \item Aligned using BWA-MEM
111 \item Genotyped using special version of GATK and Fermikit
112 \item Data available in EBI ($n=279$, PRJEB9586) and dbGAP ($n=21$, ?)
117 \begin{frame}[fragile]{Alignment Pipeline}
119 \item Aligned to the “decoy” version of the human reference
120 (hs37d5); supposedly improves alignment in misassembled regions or
122 \item PCR-free data, though they marked optical duplicates marked
126 ./htscmd bamshuf -Oun128 in.bam tmp-pre \
127 | ./htscmd bam2fq -as aln-se.fq.gz - \
129 | ./bwa mem -pt8 hs37d5.fa - \
131 | samtools view -uS - \
132 | samtools sort -@4 -m512M - out-pre
136 \begin{frame}[fragile]{Genotyping}
138 \item Reference-bias; novel variants, GATK assumes reference is more
139 likely which may not be the case. Use prior of
140 $(0.4995,0.001,0.4995)$ instead of default
141 $(0.9985,0.001,0.0005)$.
143 \item Unclear what the effect of this change is on the calling
144 \item Maybe worth thinking about?
146 \item Also used Fermikit; apparently has comparable call rates to
150 java -Xmx2g -jar GenomeAnalysisTK.jar \
151 -T UnifiedGenotyper -I srt.aln.bam \
152 -L CHR_ID -R hs37d5.fa -dcov 600 -glm SNP \
153 -out_mode EMIT_ALL_SITES -stand_call_conf 5.0 \
154 -stand_emit_conf 5.0 -inputPrior 0.0010 \
155 -inputPrior 0.4995 -D dbsnp_138.b37.vcf \
156 -o CHR_ID.vcf -A GCContent -A BaseCounts
160 \begin{frame}{Fermikit vs Platypus vs GATK}
162 \column{0.5\textwidth}
163 \includegraphics[width=\textwidth,height=0.7\textheight,keepaspectratio]{genome_diversity_paper/supplemental-014.png}
164 \column{0.5\textwidth}
165 FermiKit and Platypus call 3.17M more sites than GATK, but unclear
166 whether those are real sites or not; they go into this in much more
167 detail than I've digested yet.
171 \begin{frame}{Relatedness of Populations}
173 \column{0.7\textwidth}
174 \includegraphics[width=0.5\textwidth,height=0.8\textheight,keepaspectratio]{genome_diversity_paper/fig1a_1.png}
175 \includegraphics[width=0.5\textwidth,height=0.8\textheight,keepaspectratio]{genome_diversity_paper/fig1a_2.png}
176 \column{0.3\textwidth}
178 \item Neighbor joining tree based on pairwise divergence per nucleotide
179 \item Deepest splits are in African populations
184 \begin{frame}{PCA and Relatedness}
186 \includegraphics[width=\textwidth,height=0.6\textheight,keepaspectratio]{genome_diversity_paper/fig_ed4.png}
189 \item Greatest variation seen in the African populations (orange)
190 \item Other populations are much more similar to eachother in
192 \item Hapmap likely under-measured variation in Africa
198 \includegraphics[width=\textwidth,height=0.6\textheight,keepaspectratio]{genome_diversity_paper/fig1b.png}
201 \item Pygmy populations have lower X heterozygosity than other African populations
202 \item Seen even after removing the third of X which is subject to selection
203 \item Suggests that it's driven by demographic history, and the
204 reduced diversity is due to male-driven admixture (also in non-Africans)
208 \begin{frame}{Neanderthal Ancestry}
210 \includegraphics[width=\textwidth,height=0.6\textheight,keepaspectratio]{genome_diversity_paper/fig1c.png}
213 \item No populations studied have a higher Neanderthal ancestry than
218 \begin{frame}{Denisovan Ancestry}
220 \includegraphics[width=\textwidth,height=0.6\textheight,keepaspectratio]{genome_diversity_paper/fig1d.png}
223 \item Oceanian groups have as much as 5\% Denisovan ancestry
224 \item Eurasian differences in ancestry; some South Asians may have
225 higher Denisovan than other Eurasians
230 \begin{frame}{Variation missed by hapmap}
232 \includegraphics[width=\textwidth,height=0.8\textheight,keepaspectratio]{genome_diversity_paper/fig_ed1.png}
235 \item Hapmap is missing up to 8\% of the heterozygous sites in parts
240 \begin{frame}{Selected African cross-coalescence rate/Population Size}
242 \includegraphics[width=0.5\textwidth,height=0.7\textheight,keepaspectratio]{genome_diversity_paper/fig2a.png}
243 \includegraphics[width=0.5\textwidth,height=0.7\textheight,keepaspectratio]{genome_diversity_paper/fig2d.png}
246 \item Separation began around 200 kya for present-day hunter-gatherers
247 \item Shared ancestors as recently as 100 kya
251 \begin{frame}{Central African rainforest hunter-gatherer}
253 \includegraphics[width=0.5\textwidth,height=0.7\textheight,keepaspectratio]{genome_diversity_paper/fig2b.png}
254 \includegraphics[width=0.5\textwidth,height=0.7\textheight,keepaspectratio]{genome_diversity_paper/fig2e.png}
257 \item Within africa separation begins around 100 kya
258 \item Still intermixing between Biaka (CAR, DRC), Bantu, and Luhya (Western Kenya)
262 \begin{frame}{Ancient, non-african}
264 \includegraphics[width=0.5\textwidth,height=0.7\textheight,keepaspectratio]{genome_diversity_paper/fig2c.png}
265 \includegraphics[width=0.5\textwidth,height=0.7\textheight,keepaspectratio]{genome_diversity_paper/fig2f.png}
268 \item Separation begins around 50 kya
272 \begin{frame}{Best-fitting admixture Graph}
274 \includegraphics[width=0.5\textwidth,height=0.8\textheight,keepaspectratio]{genome_diversity_paper/fig3.png}
277 \item Present day populations have negligible ancestry from an early
278 dispersal of modern humans out of africa
283 \begin{frame}{Conclusions}
285 \item Dispersal of human populations -- little evidence of ancestry
286 from an early dispersal of modern humans
287 \item Possible acceleration in the rate of mutations among non-Africans
288 \item No evidence for species-wide selective sweeps around
289 $\approx 50$ kya (start of modern human behavior in archaeological
290 record) [Did not see that all pairs of modern humans shared a
291 common ancestor $\lt 100$ kya.
292 \item Great resource for additional variants; maybe with H3A?
293 \item Useful for my work with ancestral trees
297 \section*{References}
299 \begin{frame}[plain]{References}
301 \mbox{}\vspace{-\baselineskip}
302 \printbibliography[heading=none]